A day-ahead operational regulation method for solar district heating systems based on model predictive control

Solar district heating systems are widely used in solar-rich areas due to their centralized management and ease of maintenance. However, traditional temperature difference-based control methods do not consider the various dynamic factors affecting collector array efficiency, resulting in sub-optimal heat collection and storage. Additionally, traditional heating control methods do not consider the relationship between heat storage and demand, and district heating systems have delayed response times. This increases auxiliary heating output, leading to unstable heating stability and indoor temperature. To address these issues, this paper proposes a model predictive control (MPC)-based day-ahead operation regulation method for solar district heating systems. A sequence-to-sequence long short-term memory (Seq2seq-LSTM) prediction model is developed to forecast outdoor environmental parameters and building heating loads for the next 24 h. This prediction model is combined with the dynamic operation control model of the solar district heating systems to develop an MPC-based dayahead operation regulation model. Simulation results show that compared to rule-based control (RBC), MPC can dynamically identify optimal control points in real time, keeping the collector array within a high-efficiency operation range. Consequently, heat collection increased by 5.4 %, and the solar fraction increased by 9.1 %. MPC can balance heat storage with end-use heat demand, achieving more efficient heating by reasonably reducing the average water tank temperature. With MPC, the average water tank temperature decreased by 1.52 degrees C compared to RBC, and heat loss decreased by 3.2 %. MPC fully considers the time lag characteristics on the building side, reducing temperature fluctuations through day-ahead adjustments. The average supply-return water temperature difference with MPC decreased by 3.19 degrees C compared to RBC. The average indoor temperatures for the four types of buildings with MPC were 18.11 degrees C, 18.06 degrees C, 18.26 degrees C, and 18.12 degrees C, respectively, closer to the pre-set temperature of 18 degrees C. Finally, the total energy consumption of the system decreased by 26.5 % with MPC, including a 27.14 % reduction in auxiliary heat source energy consumption. In summary, MPC significantly improves energy efficiency, stability, and energy savings.